You should read our latest piece on Lovable’s GTM leadership team.
They had to deal with it so quickly/early and with such immense implications that this might be a paradigm shift in terms of how such AI companies approach hiring – like chess prodigies learning to think 20-30 moves ahead. While other scale-ups hire to solve their current problems, Lovable already had to think about the endgame. That’s why hires from Klaviyo/HubSpot, and that’s why 1bn ARR in 3 years.
if AI “isn’t working” for someone, it’s usually because their underlying GTM thinking is fuzzy. AI just exposes that faster.
AI doesn’t reward ambition, it rewards discipline. Most of these workflows aren’t “advanced.” They’re just unsexy. Feed real data. Add constraints. Keep humans in the loop. Repeat until it compounds.
What clicked for me is that the advantage isn’t the AI... it’s the architecture. Instruction stacks. Persistent context. Feedback loops. That’s why two teams with the same tools get wildly different outcomes.
Source: what truth do you already have? (transcripts, lost reasons, objections)
System: 1 workflow, not 10 tools
Ship: run it weekly like payroll
Score: did it save time or make money?
Iterate: tighten prompts + inputs, kill fluff
Because speed isn’t the win. Feedback loops are. The teams compounding are the ones turning AI outputs into tracked decisions (what changed, why, result).
Question for you: if someone is in the “53% no impact” bucket, what’s the single starter workflow you’d mandate for 30 days
Got referred here by Maja and it wow what a knowledge journey this article is. Will definitely use "Content ideation" and "Lead Enrichment" workflows.
Thanks for the great article.
so glad you liked it! Thanks so much for your feedback! New newsletter coming out every Friday - hope you'll like it here
You should read our latest piece on Lovable’s GTM leadership team.
They had to deal with it so quickly/early and with such immense implications that this might be a paradigm shift in terms of how such AI companies approach hiring – like chess prodigies learning to think 20-30 moves ahead. While other scale-ups hire to solve their current problems, Lovable already had to think about the endgame. That’s why hires from Klaviyo/HubSpot, and that’s why 1bn ARR in 3 years.
https://thebigbyte.substack.com/p/how-lovable-is-building-its-gtm-leadership
A refreshingly practical breakdown of where AI is truly driving GTM results, with concrete workflows that cut through the hype and focus on execution
if AI “isn’t working” for someone, it’s usually because their underlying GTM thinking is fuzzy. AI just exposes that faster.
AI doesn’t reward ambition, it rewards discipline. Most of these workflows aren’t “advanced.” They’re just unsexy. Feed real data. Add constraints. Keep humans in the loop. Repeat until it compounds.
What clicked for me is that the advantage isn’t the AI... it’s the architecture. Instruction stacks. Persistent context. Feedback loops. That’s why two teams with the same tools get wildly different outcomes.
I think a good flow should be:
Signal → Source → System → Ship → Score → Iterate
Signal: what metric moves? (meetings booked / pipeline / cycle time)
Source: what truth do you already have? (transcripts, lost reasons, objections)
System: 1 workflow, not 10 tools
Ship: run it weekly like payroll
Score: did it save time or make money?
Iterate: tighten prompts + inputs, kill fluff
Because speed isn’t the win. Feedback loops are. The teams compounding are the ones turning AI outputs into tracked decisions (what changed, why, result).
Question for you: if someone is in the “53% no impact” bucket, what’s the single starter workflow you’d mandate for 30 days